Note
Nebari is evolving — this repository is now called nebari-classic.
Over the years of building and maintaining Nebari, we accumulated a long list of patterns, tooling, and architectural decisions we wanted to apply but were constrained by the current architecture. Rather than continue layering on that foundation, we decided to pause feature development on what we now call nebari-classic and invest in a new architecture organized around the expertise we've built:
- nebari-infrastructure-core (NIC) — a composable, modular approach to infrastructure management.
- A set of component layer "packs" that allow software to be easily integrated on top of NIC through GitOps: including Nebi (A newer and more resilient replacement for conda-store), Dask, and the data science pack (JhubApps, Jupyter* and more). All that together covers the user-facing functionality nebari-classic provides today. Browse them under Software Packs on the docs site, or search the nebari-dev org for "pack" repositories.
See the NIC design documents in that repository for the full reasoning behind this transition.
What this means for nebari-classic. We will continue to ship critical bug fixes, security and vulnerability patches, and dependency updates on this repository. Though at a slower cadence than during active feature development. We expect to keep the maintenance window open until the new architecture is stable enough and well-tested so that operators have had time to migrate, likely through the end of 2026.
Nebari as a whole is very much still being built — you are more than welcome to join us at the community meetings and help shape what comes next.
| Information | Links |
|---|---|
| Project |     |
| Community |    |
Nebari was an open source data platform for building and maintaining cost-effective, scalable compute platforms on HPC or Kubernetes with minimal DevOps overhead.
It used Terraform, Helm, and GitHub Actions to deploy JupyterHub with Dask Gateway on Kubernetes across AWS, GCP, and Azure.